DEAD FUEL MOISTURE CONTENT ESTIMATION USING REMOTE SENSING

Abstract

One of the critical parameters in wildfire behavior is the dead fuel moisture content (DFMC). DFMC is affected from environmental factors and the vegetation characteristics, thus it variesacross the landscape. Previous research showed that remote sensing reflectance data can assist the spatial estimation of DFMC. The aim of this paper is to evaluate the Landsat 8 in retrieving the DFMC in a complex Mediterranean ecosystem. The Normalized Difference Vegetation Index (NDVI) and the top of atmosphere brightness temperature are correlated with the 10-h fuel moisture content and the surface temperature recorded from Remote
Automatic Weather Stations (RAWS). Training data are collected from the year 2015 and the validation is applied to year 2016. According to the literature, the DFMC was correlated with the ratio of NDVI/LST, however, our results were not satisfactory producing low r2 coefficients. New models were developed based on the DFMC and the brightness temperature (BT) which resulted to r2 values up to 0.733. The validation with new data confirmed that the top of atmosphere brightness temperature retrieved from Landsat 8, can be used as an input to estimate the spatial distribution of DFMC. The process was fully automated, i.e. from data ordering to map preparation, and is ready to continually provide the wildfire managers and firefighting personnel confronting wildfires with the DFMC maps.